We propose efficient and low-complexity multiuser detection (MUD) algorithms for Gaussian multiple access channel (G-MAC) for short-packet transmission in massive machine type communications. To do so, we first formulate the G-MAC MUD problem as a sparse signal recovery problem and obtain the exact and approximate joint prior distribution of the sparse vector to be recovered. Then, we employ the Bayesian approximate message passing (AMP) algorithms with the optimal separable and non-separable minimum mean squared error (MMSE) denoisers for soft decoding of the sparse vector. The effectiveness of the proposed MUD algorithms for a large number of devices is supported by simulation results. For packets of 8 information bits, while the state-of-the-art AMP with soft-threshold denoising achieves 8/100 of the upper bound at Eb/N0 = 4 dB, the proposed algorithms reach 4/7 and 1/2 of the upper bound.

翻译：针对大规模机器类通信中的短数据包传输，我们为高斯多址接入信道提出了高效低复杂度的多用户检测算法。首先将高斯多址信道多用户检测问题建模为稀疏信号恢复问题，并推导待恢复稀疏向量的精确与近似联合先验分布。随后采用基于贝叶斯近似消息传递的算法，结合最优可分离与不可分离的最小均方误差去噪器，实现稀疏向量的软解码。仿真结果验证了所提算法在大规模设备场景下的有效性。当传输8比特信息数据包、信噪比为4 dB时，采用软阈值去噪的现有最优近似消息传递算法可达到理论上限的8/100，而本算法分别达到4/7和1/2的理论上限。